Kiwi Co-founder Narrative: Why is it difficult for encryption consumption applications to achieve product-market fit?

Author: Mac Budkowski, Deep Tide TechFlow

Original Title: Kiwi Co-founder: Those Things About Building Consumer Crypto Applications

In the past few years, I have been dealing with cryptocurrency founders. I have interviewed them on my podcast, interacted with them at conferences, and became one of them when creating a startup called Kiwi focused on Ethereum hacking news.

Since I left Kiwi in June, I have had some time to reflect and think back. I do believe that consumer-facing crypto projects are quite different from regular consumer startups.

The reasons are as follows.

Finding Water Sources

We all know that the goal of startups is to find the Product-Market Fit (PMF), which is a magical combination—the perfect blend of product features and distribution channels that allows companies to grow rapidly.

Finding PMF is like using a magic wand to search for a water source.

When the water source is near, the magic wand vibrates. But the problem is that when you are eager to find the water source, it's easy to mistake the vibration of the magic wand for your anxious hands shaking.

In the field of cryptocurrency, this search becomes even more challenging, as the environment is filled with false signals that can cause your “magic wand” to vibrate incorrectly. You may waste a lot of time moving in one direction, only to eventually discover that there is no water source there at all.

Let's start with an example.

When an income of 1 million dollars fails to provide a clear signal

You may have heard of Farcaster, which is a decentralized social application. This year, they launched Farcaster Pro, a paid service similar to Twitter Premium. However, when it first launched, Farcaster Pro did not offer many advanced features.

Surprisingly, Farcaster Pro generated over $1 million in revenue within the first 24 hours of its launch! And this is in the consumer-facing crypto space. It seems like a moment worth popping the champagne, right?

But… if you analyze it deeply, you will find a strange phenomenon: nearly half of the users who purchased a Farcaster Pro subscription have fewer than 100 followers. It is important to note that Farcaster Pro is designed for “experienced users” (which is also the reason it is named “Pro”).

Why would these users with a low follower count spend $120 on a subscription?

Some users may purchase subscriptions to obtain exclusive badges to make their profiles more prominent and attract more followers. However, more users buy subscriptions for speculation. Moreover, this speculation has indeed paid off: the first 10,000 users who purchased Farcaster Pro received an airdrop reward worth $600 the next day, achieving a 5x return. A 5x return within 24 hours is certainly impressive.

Interestingly, the aforementioned post was not written by an airdrop user (a user specifically chasing airdrop rewards), but by Chris. He is a long-term senior user of Farcaster and an early supporter.

This indicates that there are at least two types of people who purchased Farcaster Pro: speculators and believers.

This also reveals the first reason why signals are distorted in the crypto space: when you are looking for PMF, you want market validation from users who genuinely want to use the product. However, as we see here, you can make a million dollars in 24 hours and still be uncertain whether you have gained market recognition.

This is not the fault of the Farcaster team. They do want to create a real consumer product, and the tokens for this airdrop were not initiated by them. But this is the reality of the cryptocurrency industry, and it is the environment we all find ourselves in. This environment makes it difficult to determine whether we are truly “building what people want,” as Paul Graham said.

And this is just the tip of the iceberg.

Airdrop hunters are not always glamorous

When we launched Kiwi, users needed to purchase an NFT worth $10 - the Kiwi Pass, to fully use the product. In the first three months, over 100 people made purchases, including some well-known users like Fred Wilson. This was entirely organic growth, with no marketing push.

At that time, I thought to myself:

“Have we had paying users from day one in a bear market? Doesn't that mean we've found some breakthrough point?”

However, we soon realized that things were not that simple.

We found that some people buy Kiwi Pass because they are friends and want to support us; some do it because they agree with our commitment to decentralization and open source principles; and some pay this 10 dollars because using “magical internet currency” feels easy, so why not give it a try?

This means that, although there are no airdrop hunters in our user base, only a portion of the paying users actually want to use our products.

If you think about it carefully, this is really crazy.

If you ask web2 founders, they might ask about vanity metrics like “Twitter followers” or “page views.” But I don't think anyone would mention “revenue,” as that is the most fundamental entrepreneurial metric.

If you ask Web2 entrepreneurs what vanity metrics are, they might say “Twitter followers” or “page views.” But hardly anyone would say “revenue,” because revenue is the most fundamental metric for startups. However, in the crypto space, this is less clear. Even if you see revenue growth, you cannot determine whether you are moving in the right direction—this may be due to speculators or well-meaning supporters. Thus, a seemingly reliable “PMF signal” may actually mislead you.

However, the market structure will not help solve this problem.

Why is the cryptocurrency consumer market so difficult

If you include the mainnet and all L2s, Ethereum has about 45 million active addresses each month.

If we assume that each person has only one address (which is not the case in reality), this means that Ethereum's market penetration rate globally is less than 1%.

Therefore, to use Geoffrey Moore's classic terminology – most users (in the West, as the situation differs in Turkey or Argentina) are innovators and early adopters.

These users are very interested in the cryptocurrency space, and even if the user experience (UX) is poor, they are willing to test your product. This sounds good, but this double-edged sword also brings problems. Because these users are willing to try new products, your usage curve may spike in one week, but the next week they might turn to the next eye-catching thing.

This is the first question. The stickiness of these users is far lower than that of “regular users.”

In addition, the cryptocurrency market is very diverse.

Taking Kiwi as an example, our user base includes developers, founders, artists, venture capitalists, and researchers, all of whom have very diverse expectations for content. Some people want to see more technical content, some want to see more artworks, and others want to see more articles. This makes it challenging for us to meet everyone's needs without losing focus.

The duality of the Ethereum community

This diversity applies not only to Kiwi but also to the entire crypto space. Here, you will also encounter speculators, philosophers, cryptographers, traditional finance professionals, and more. This means that this already small market is further divided into multiple subgroups, which are dispersed across different L1 and L2 networks.

This is a challenge for growth and product design:

• Growth Issues: With so few users, how do you find and attract them?
• Design issue: How to filter feedback from multiple user groups? The needs of these groups often contradict each other, and their reasons for entering the market vary.

In a large market like coffee, you have enough space to accommodate many niches. For example, there is a multimillion-dollar business run by veterans that specializes in providing coffee for people who love America.

However, choosing a niche that is too obscure in the crypto market may lead to an inability to establish a sustainable business. Even if people continue to use your product, you may not be able to find enough users to cover the costs.

As pmarca said:

“In a great market—a market with a large number of potential customers—the market will actively pull the products of startups.”

However, in the field of cryptocurrency, apart from stablecoins, speculative activities, and a few niche applications like prediction markets or collectibles, we do not yet have a “large number of potential customers.”

Therefore, apart from those users who distort signals and those chasing the “next hot thing,” our market size is also very small. And if you want to find PMF, this is crucial.

PMF is not just growth

Most people equate PMF with growth.

The app suddenly gained 100,000 downloads in a week? PMF!

All friends start using your product? PMF!

The agreement generated a trading volume of 10 million USD within a month? PMF!

But in reality, PMF is not just about growth; it also concerns user retention.

If customers come to your store, look around and leave, is that considered a successful store? Of course not. However, we often assume that applications with high download numbers must have product-market fit (PMF), forgetting counterexamples like Clubhouse, which quickly declined after a brief period of popularity.

There are many similar stories. The Twitch team created 16 million downloads within 4 months during the early development of their app. And—according to its founder—they still had not achieved PMF.

If you want to hear this story, please go to 7:30 in the video below:

Original video link: Play video

They do not have PMF because their retention rate is too low, and their app is basically just burning through the market. Users come to their store, then leave, and never come back.

In the crypto space, we are no strangers to such high growth/low retention stories, as we have all seen airdrop hunters. But the problem does not stop there.

But things are far from over.

There is also the influence of market trends. If you built an NFT marketplace in 2020, you would have garnered a lot of attention; but if you tried to do the same thing in 2025, the outcome could be completely different. These trends come and go, and you might think you succeeded, but when the trend fades, your product may disappear as well.

How Tokens Affect User Attention

In addition, there are two key factors at play.

The cryptocurrency space is rife with dozens of zombie projects that siphon off the attention and funds of retail investors. Since announcing a project’s death is considered taboo, founders have all “exited the community,” while their users continue to linger in the project’s Discord channels, believing that their tokens, NFTs, or points will one day yield returns.

This is a brutal zero-sum game. If users still believe that Cardano will win, they won't take the time to understand Ethereum or your decentralized application (dapp), because they are already surrounded by the promotional content of their preferred network.

Another factor is that the price of cryptocurrency tokens may affect user retention.

For example, during the Base memecoin frenzy in 2024, we noticed a decline in the usage of Kiwi. Why is that?

When users can choose to spend 15 minutes reading an article about decentralization or spend 15 minutes looking for the next coin that can 100x, even those who originally came to the crypto space for legitimate reasons will choose the latter.

There was a time when artificial intelligence + memecoins attracted people's attention away from everything else.

What should you do in the face of such a challenging market?

In summary: We are facing speculators, users who support you but do not use the product, a small and diverse market, challenges in user retention, and the competition for user attention has become more difficult due to token prices.

The original intention of writing this article is to describe the reality I see, rather than to discourage entrepreneurs from innovating in the crypto consumer space.

So I owe you some solutions. Although these methods haven't made Kiwi the “next big hit” (at least so far!), they have indeed made our work a bit easier.

Question 1: Speculators

Do not create distortion.

If you launch a points or referral program, it will be difficult to determine whether people genuinely like your product or are just expecting to receive some monetary rewards. Therefore, if you are looking for early signals, do not do this. It is a trap!

We once launched a promotional plan on Kiwi, but the results were not ideal.

Until one day, we saw a large influx of users. Upon deeper investigation, we found that most of the users were from South Korea.

We soon learned that a Korean influencer shared his referral link and claimed that users would receive airdrop rewards. We did indeed gain many registrations, but there were very few long-term users. Since we were looking for signals rather than an extra $200 in income, we decisively terminated the plan.

By the way, not creating distortion is one of the reasons why Kiwi Pass was non-transferable in the past—we do not want people to speculate on it, as this would add more noise.

Reward good user behavior

At Kiwi, we reward users with Karma. Just like on Reddit, the more upvotes your posts get, the more Karma you earn.

Users can see the most active curators on the leaderboard, and we will also commend them in our weekly newsletter, recognizing their efforts. This is a positive social reward mechanism that encourages users to be good users rather than speculators.

In addition, we have clear guidelines outlining the types of content we want to see. If someone shares low-quality junk content, we can easily kick them out because their links do not meet the guidelines.

If we cannot make a judgment, we will check the on-chain history to determine whether they are ordinary users or non-compliant speculators. Although not every product can add guidelines, there are many ways to reward good user behavior.

Question 2: People do not use the product

Communicate with users

When you mint a Kiwi Pass, we will invite you to join our Telegram group chat. This is a major advantage in the crypto space - most people use the same messaging app. Getting casual feedback through Telegram is 10 times easier than through email.

In the Kiwi group chat, we share product updates, discuss crypto news, gather feedback on feature ideas, and more. This also allows us to reach out to the most active users via private messages.

We have communicated with hundreds of users, whether through Telegram or video chats, to understand how to make the application more attractive. Sometimes I feel like a missionary, asking users if they are willing to talk about “Jesus,” but it is indeed worth it.

Segmented feedback

If you receive feedback, it is crucial to understand its source.

Social applications have a “90/9/1 rule”, where 90% of users only read content, 9% interact with the content, and 1% are high-frequency content publishers. This situation is similar in Kiwi.

We collect feedback from different types of users and then develop features to address current bottleneck issues.

When content is insufficient, we have developed a Chrome extension for creators, allowing users to submit links with one click. When comments are lacking, we focus on developing features for contributors, such as improving the comment editor. When we want to attract more readers, we focus on developing features for divers, such as optimizing app load times, Feed algorithms, and so on.

Understanding which group of users you are talking to can be very helpful.

Communicate with churned users

When we find that certain users have become inactive, we will contact them to try to understand what has happened.

Users often politely say they are “busy”, but upon deeper probing, some become outspoken.

Some people say they “have been logged out multiple times and are tired of logging back in,” or “their links haven't received enough likes,” or even “they feel that reading our links takes more time than scrolling through Twitter's feed.” These are all pieces of information we can refer to.

Question 3: Small and Diverse Market Size

Looking for high ARPU markets

If you had to choose which “Winnie the Pooh” to serve, would you choose the regular version or the one in a suit?

If in doubt, choose the version in a suit, as he has greater purchasing power. Even if there are not many users, you can still make a lot of money through a high Average Revenue Per User (ARPU).

Last month, Aave collected over $60 million in fees on Ethereum, with approximately 25,000 monthly active users. This means that each user incurs fees of over $2,400 per month. This is due to some “whales” borrowing and lending millions of dollars. Of course, not all fees go into Aave's treasury, but it is still an astounding figure.

So, you don't need a lot of users; you need the right users.

Similarly, Blur focuses on a very niche market of professional traders and has still achieved excellent results when everyone thought the NFT market was dead.

Expand the market

If your niche market is too small, but you believe it has potential, you can choose to actively expand the market size or wait for the market to grow naturally, just like Master Splinter in the Teenage Mutant Ninja Turtles.

In 2022, the privacy sector on Ethereum was a small market. But by 2025, it had grown large enough for Raligun to generate $5 million in revenue. And because they have been around for 3 years, they became the preferred solution for users as the market grew. The same goes for OpenSea, which established its market position before NFTs became the “next big thing.”

Crossing the chasm

A more aggressive market expansion approach is to cross the chasm and attract ordinary users.

NBA Top Shot is an example. They made collectibles more attractive to ordinary users by partnering with the NBA and provided a simple way to purchase their first NFT. This strategy worked!

Another example is Polymarket. They brought prediction markets into the mainstream—partly due to their top-notch operations on social media, partly because of their successful predictions of the U.S. presidential election, and partly due to their creative innovations in betting.

If you decide to take this path, I recommend that you focus on niche markets with 100,000 to 1 million users. You can read more in my article, which critiques the efforts to bring the next billion users to Ethereum.

Question Four: Difficult to Retain

Create a “toothbrush-shaped” product

Google co-founder Larry Page once stated that he is looking for a “toothbrush” product - a product that is used once or twice a day and improves life.

If you focus on the core functionality of the product and design it to be used daily, it will bring two benefits:

First, the positioning of the product will become clear. For example, Kiwi's core function is “discovering quality Ethereum content.” When there is excellent content in our content stream, usage will increase because that is exactly what users need. Focusing on a core function also makes user feedback more concentrated—you can have 20 conversations around one theme, rather than having 20 conversations around 20 different themes.

Secondly, when you focus on daily issues, you can receive feedback every day, allowing for more frequent adjustments in direction. If you focus on monthly issues, the adjustment cycle will be longer.

Of course, this does not mean that excellent products cannot be built around weekly or monthly issues. For example, most people do not stake their ETH every day, but Lido is still very successful. However, finding daily interaction opportunities may help to gather more signals.

The simplification of this “toothbrush-shaped” product has another benefit: it can reduce time wasted on non-priorities. For example, one day we noticed that Kiwi's personal homepage was loading very slowly. We could have spent time fixing it, but since users did not come to Kiwi for a complicated personal homepage, we directly removed many features from the homepage to make it faster, while focusing on improving the core functionalities.

Enhance value realization time

If I cook instant noodles, my value realization time is about 1 minute—I just need to pour some hot water to have a meal. If I order in a restaurant, my value realization time is about 30 minutes, but the dishes are more delicious, so the value I get is higher than that of the noodles.

For us, our value comes from content. But the problem is that most of our content is long, making it more like a restaurant—users need to spend a lot of time to get a quality experience.

But we found that on Hacker News, many users do not click on links; they just read the discussions below, which is a quicker way to “consume” content. This prompted us to work on making comments more popular, such as adding comment previews, emoji reactions, and a more mobile-friendly way to write comments.

We also tried more “instant” content, like memes or charts, which may have lower value, but users can access them immediately.

Focus on the high alignment between products and users.

If your mom downloaded a DeFi aggregator app, the product-user fit would be very low; she would feel like a confused bird at a conference. But if you attract a DeFi veteran who trades 5 times a day, the fit will be very high. The higher the fit, the higher the potential user retention rate.

When we attended the Ethereum conference, many people were interested in Kiwi and immediately downloaded the app, with some becoming long-term users. This is because the product-user fit is very high, which is why sponsorship fees for top conferences can exceed $10,000 — all products want to attract such a user base.

Create a nice mobile application

I know that the Ethereum ecosystem is filled with desktop-first users.

Yes, tools like Figma and Cursor do not need to worry about mobile applications, but consumer products do. So, please create an excellent mobile application. We have tried PWA (Progressive Web Apps), and while it performed well in the short term, it is not ideal in the long run due to limited analytical capabilities, poor push notification effectiveness, and it tends to crash more easily because users are using different mobile browsers.

Therefore, you only need to get the iOS or Android application.

Question 5: Tokens and failed projects capture user attention

Explore “small channels”

Andrew Chen proposed a concept called “small channels.” If your application has 100 users and gains 10 new users in a week, then your application’s growth rate reaches 10%, which is great.

Therefore, when your product is at a smaller scale, you can focus on some smaller, high-signal channels, such as subreddits, forums, niche podcasts, etc., rather than trying to go viral on Twitter. This is also how ZORA acquired its first users—they directly messaged artists on Instagram.

When your product is small in scale, you can even explain your product to people one by one, just like the person in the meme.

We participated in professional podcasts such as Dev N Tell with Kiwi, increasing exposure for the Gitcoin project and attracting users one by one, among other things. Some methods were not very effective, but some performed well, such as participating in the ENS ecosystem conference call, where 50% of the attendees registered for Kiwi immediately.

It is precisely because of this high level of focus that you can find people who genuinely care about technology rather than token prices. Moreover, their retention rates are also higher.

Attend Ethereum conference

One benefit of the Ethereum conference is that you can hardly hear anyone discussing prices. Attendees are there for the technology.

This is also a very reliable way to collect feedback. For example, we learned how to improve the user onboarding process and why the mobile Ethereum wallet isn't working properly through this method. We were also able to observe the application's interaction performance across different browsers, operating systems, and wallets.

By the way, this is how Railgun is learning to improve its documentation.

You will never know the result before the release

This is my last suggestion.

All of these suggestions are aimed at early-stage projects.

Early-stage projects refer to: “You have just launched a product and are looking for signals to determine if you're heading in the right direction.” If your goal is to enhance data to raise the next round of funding, or if you have already found PMF and wish to scale up, it is better to leverage some noise for greater growth.

When it comes to early projects, my last piece of advice, though obvious, must be said: none of these ideas will work until you get the product out there. Therefore, releasing early is always the better choice.

When Tim released Kiwi, you had to download the Kiwi GitHub repository, set up the node, and submit the link via CLI. From the perspective of consumer products, this is overly complicated. But it was enough to pique the interest of some users (including me), and that's how we began collaborating with Tim.

Finding signals is not an easy task.

In my opinion, entrepreneurs are like living in a 24/7 casino, playing a game. In this game, if you combine the right features and distribution strategies in the correct order and hit “Publish”, you might win the jackpot. But even experienced founders and venture capitalists find it difficult to judge whether the combination is correct before hitting the “Publish” button.

It is more like art than science. In the field of art, you can cultivate your taste and perception of trends, but that does not mean you can always predict which song will become the champion of the Billboard Hot 100.

For example, the biggest hit song of the band Black Sabbath, “Paranoid” — which they chose as the closing song for the band's final performance — was actually written in just 20 minutes and was originally just a filler track on the album. Quentin Tarantino and his director friends believed that “Pulp Fiction” would be a dismal failure. And “The Shawshank Redemption,” which ranks first on IMDB, actually had a poor box office performance upon its release, only gradually establishing its legendary status after becoming a hit in the home video rental market.

Entrepreneurship is the same - it's hard to predict what will succeed. After all, there are many commonalities between consumer products and consumer culture.

So, please don't be discouraged by all the challenges I've mentioned, go ahead and launch your product, good luck!